Assessment and development of membrane materials and chemistries for the iron-chromium redox flow battery

Abstract

The mitigation of climate change demands large-scale energy storage solutions (LSESSs) that can integrate electrical grids with renewable energy sources, such as solar and wind. Flow batteries (FBs) have been identified as promising LSESSs due to their high safety and scalable capacity, which is coupled to their electrolyte volumes. While the vanadium flow battery (VFB) is considered the most developed FB, the high cost of electrolyte remains the most significant barrier when competing with lithiumion (Li-ion) batteries. The electrolyte of the iron-chromium flow battery (ICFB) is made from cheaper and more abundant metals, which would enable a more cost-effective up scaling. Additionally, the large ferrochrome ore reserves of South Africa, coupled with its low population density and high annual solar irradiance, gives the technology a unique opportunity to benefit the country by storing renewable solar energy, thereby reducing CO2 and pollutant emissions by replacing coal plants. To feasibly reach grid-scale development, flow batteries need to be low maintenance, safe, highly cost-effective, environmentally benign, while having recyclable active materials, long cycling life and high efficiencies. The high cost of the benchmark membrane used in the ICFB (perfluorinated sulfonic acid (PFSA)) is the biggest challenge towards commercialisation of the technology, however, there is limited literature focussing on alternative suitable membrane materials. While these PFSA membranes are highly chemically stable, they also exhibit a low metallic-ion selectivity with severe electro-osmotic crossover of iron-chrome electrolyte volumes. To address this research gap, various membrane materials and chemistries were assessed and developed in this study. After construction and optimisation of a lab-scale ICFB test station, the first materials tested were simple and low-cost commercial hydrocarbon-based microporous separators (MPSs) that had been developed for the lithium-ion and lead-acid battery industries. Despite low air-permeabilities and relatively equalised differential pressures, most of the tested MPSs displayed high crossover rates resulting in reduced energy efficiency (EE) values and self-discharge times (4.3% and 40.5% below the benchmark cation exchange Nafion-212 (N-212) membrane, respectively), where 6 of the 10 MPSs were suitable for short-term cycling (10 cycles). The dynamic behaviour of the asymmetrical electrolyte viscosities, linked to the state of charge, led to changes in the differential pressures across MPSs which worsened convection. Despite pulse dampening and asymmetrical pumping, which reduced the convection and capacity decay, MPSs should be further optimised specifically for the ICFB. Further research should focus on MPS thickness and wettability, since they were shown to have the largest impact on the performance of an ICFB. Since anion exchange membranes (AEMs) have the potential for a high cation selectivity with no published successful cycling in an ICFB to date, a range of AEMs were manufactured and tested. Most of the AEMs failed to discharge the electrolyte due to membrane fouling by ferric chlorides, where adding sulfates (as sulfuric acid) only worsened the measured resistance due to sluggish anion migration. However, one cross-linked AEM consisting of m-polybenzimidazole (m-PBI) and phosphonated poly(pentafluorostyrene) was able to charge and discharge the ICFB electrolyte without electro-osmosis, yielding a 4.9% higher 30-cycle average coulombic efficiency (CE = 96.7%) and 1.3% higher EE (76.1%) than the N-212 cation exchange membrane (CEM). The successful application of mPBI was attributed to the high degree of swelling, due to the protonation of the imidazolium groups, that likely enabled proton migration through enlarged molecular spaces and electrolyte channels. Finally, a wide variety (nanofibre reinforced, phosphonated, sulfonated, blends, ionically cross-linked and ionically-covalently cross-linked blends) of cation exchange membranes (CEMs) were manufactured and tested in the ICFB. Initial screening results showed an inherent incompatibility between phosphonic acid-based ionomers and the ICFB electrolyte. Accordingly, sulfonated ionomers were developed further. A low-cost and highly sulfonated poly(ether ether ketone) SPEEK was cross-linked with a diphenylether-containing PBI (OPBI) and optimised for the ICFB in terms of conductivity and selectivity by varying the acid-base blend ratios. A 55 µm CEM with a SPEEK-95 to OPBI blend ratio of 89:11 obtained a 3.3% lower EE than N-212 and 1.7% higher CE, while reducing the benchmark 30-cycle electrolyte imbalance levels from 33% to 4%. A novel sulfonated ionomer, SFS, was ionically and covalently ([1,1'biphenyl]-4,4'-dithiol) cross-linked with OPBI and optimised, yielding an IEC of 1.54 mmol g−1, outperforming the benchmark N-212 with an EE of 1.4% (76.2%) with no electro-osmotic crossover. Various membrane types (MPSs, AEMs and CEMs) of cost-effective materials that have not previously been considered for the ICFB were sourced and manufactured in this study. The ICFB feasibility of these PFSA alternative materials, including hydrocarbon-based separators, polybenzimidazoles and various other aromatic polymers, were demonstrated on lab-scale. The AA900 MPS, the MIG-15 AEM and three OPBI containing CEMs (SPEEK-OPBI 89:11, SFS-OPBI 84:16 and SFS-OPBI (I+C)) all had comparable efficiencies, but significantly lower electro-osmosis, than N-212. Combining the different advantages of the ion-selective AEMs, conductive CEMs and low-cost porous hydrocarbon separators could further improve their ICFB performance. Longer term studies (multiple years of cycling) should also be considered to further validate the promising alternative materials. This would however require a capacity rebalancer, as well as the development of an ICFB electrolyte containing anionic ligands, alternative to Cl− and H2O, to ensure a sustained discharge capacity and Cr couple redox activity.